ES2317956T3 - COMPENSATION OF THE TIMING DEVIATION IN WIRELESS SYSTEMS BASED ON PACKAGES. - Google Patents

Abstract

A radio receiver system ( 30 ) comprises a radio receiver ( 41 ), a receiver sample clock ( 60 ), which is used for sampling a modulated base-band signal; and a timing correction unit ( 100 ). The timing correction unit ( 100 ) performs, in the frequency domain, a timing drift compensation between a transmitter sample clock ( 66 ) and the receiver sample clock ( 60 ). In one example context of implementation, the plural modulated radio frequency carriers have been modulated using Orthogonal Frequency Division Multiplexing (OFDM).

Description

Timing Deviation Compensation in wireless systems based on packages.

Field of the Invention

This invention belongs to an apparatus for determine and / or compensate for a time deviation between clocks of sample of a transmitter and a receiver along with a transmission of several modulated signal carriers on an air interface or radio.

Background of the invention

Various methods for signal transmission digital, such as digital video broadcasting (DVB) signals (digital video broadcast) and digital audio broadcasting (DAB) (digital audio broadcast), are known. A typical method used for such transmissions is the orthogonal frequency method division multiplexing (OFDM) orthogonal frequencies) in which a plurality of carriers of Modulated signal is used to emit the signals. Schemes of Multi-carrier modulation such as OFDM is typically used. in systems in which their time dispersion is much longer than the bit duration used. In the schemes of Multi-carrier modulation, the modulated signal carriers are sampled before being transposed in the domain of the frequency by means of a fast Fourier transformation (FFT) (Fast Fourier Transformation) for signal separation.

The Orthogonal Frecuency Division Multiplexing (OFDM) (Orthogonal Frequency Division Multiplexing) is will also use in wireless local area networks (WLAN) Wireless Local) in the 5 GHz band as specified in Europe, the U. S. and Japan. The European WLAN standard is High Performance Radio Local Area Network type 2 (HIPERLAN / 2) (Network of Type 2) High Performance Local Radio Area, which is being developed by the ETSI project (European Telecommunication Standard Institute - Telecommunications Standards Institute European) BRAN (broadband radio access network) broadband radio). North American standards are expected and the Japanese have very similar physical layers.

The physical layer of HIPERLAN / 2 is described in the ETSI TS101 475 VI. 2.2 (2001-02), Broadband Radio Access Networks (BRAN) (Broadband Radio Access Networks); HIPERLAN Type 2; Physical (PHY) layer, ETSI document. He HIPERLAN / 2 air interface is based on multiple access by Duplex time division and dynamic time division. All the data units in each transport channel that are transmitted through the physical layer of HIPERLAN / 2 are bursts

The frame structure 18 and a burst 20 of example are shown in Fig. 1. The frame includes a channel of broadcast control, a frame control channel, downlink traffic, Upload traffic, and random access channels. As shown in Fig. 1, each burst 20 comprises a preamble 22 and one or several PDUs (Unit Directed by Preamble) 24_ {1}, 24_ {2}, 24_ {3}. 24_ {n}. Other details of the preamble 22 and the PDU's 24_ {1}, 24_ {2}, ... are illustrated in Fig. 2. Preamble 22 includes, for example, a cyclic prefix 22_ {1} and two symbols of training 22_ {2}, 22_ {3}. Each PDU 24 is a data unit multi-byte that comes from the medium access control layer (MAC) (medium access control).

The physical layer (PHY) converts the PDU to one or more OFDM symbols, depending on the number of symbols in the mode of link adaptation. While Fig. 1 shows a PDU 24 of example comprising several OFDM symbols, for the sake of simplicity Fig. 2 shows its PDUs comprising a prefix 25 cyclic and a data symbol 26. For example, in Fig. 2 the PDU 24, has a cyclic prefix 25, and the data symbol 26_ {1}.

OFDM modulation is carried out with an IFFT (Reverse Fast Fourier Transform) (Fast Fourier Transform Reverse) of 64 points at a sample rate of 20 MHz, which provides a subcarrier separation of 312.5 kHz and a symbol duration of 3.2 \ mus. The 25 cyclic prefix set in front of each symbol 26 is 800 ns, providing a length 4 \ mus total symbol Of the 64 subcarriers, only used 52 subcarriers, of which 48 carry data and 4 are pilots.

The physical layer provides several modes of Link adaptation to adapt to various channel conditions. Each mode comprises a combination of a modulation scheme of subcarrier and a correction coding rate of subsequent error The primary modulation schemes are BPSK, QPSK, 16QAM and 64QAM. The primary coding rates are 1/2, 9/16, 3/4.

In order to perform a demodulation consistent, a receiver must synchronize with the transmitter both in time as in frequency. Due to frequency differences between transmitters and receivers in such systems, the carriers Demodulated signal may present frequency deviations. An estimation of the channel must also be made.

The preambles 22 in the data stream facilitate, for example, an analysis of frequency deviation of the received signals and frequency synchronization. In the case of OFDM, the two symbols of OFDM symbols 22_ {2}, 22_ {3}, identical also called C64, are inserted between the cyclic prefix 21 (C32) and the actual data stream, for example, PDUs 24. This so-called preamble C shown in Fig. 2 is use, for example, for a channel estimate in the process of demodulation of multi-carrier signals. So, it takes place an estimate of the channel at the beginning of the burst with the help of Preamble, which contains the training symbols.

In the HIPERLAN / 2 the 20 MHz clock shows both the transmitter and the receiver move freely with a relative accuracy of +/- 20 parts per million (ppm). Worst case scenario is therefore a deviation of +/- 40 ppm between Sample clock frequencies in the transmitter compared to the receptor. This potential deviation of +/- 40 ppm causes a timing error between the transmitter and the receiver that increases over time, that is, a timing deviation. This timing deviation may cause a disturbance. significant in consistent demodulation at the end of a long burst, even if the receiver is synchronized at the beginning of the burst.

The timing deviation causes some unwanted effects, called: (1) phase rotation in the domain of the frequency; (2) interference between symbols; and (3) some loss of orthogonality between subcarriers. It is not considered that the last two effects influence demodulation. But nevertheless, As illustrated below, the phase rotation effect can be meaningful

The Fourier transform ratio of the Equation 1 indicates that a time offset T_ {off} of a function f (t) in the time domain provides a factor of linear phase \ omegaT_ {off} of the corresponding transform of Fourier, compared to the Fourier transform F (w) of the original f (t) function.

       \ vskip1.000000 \ baselineskip
    

one

       \ vskip1.000000 \ baselineskip
    

Considering a burst with the length of a 2 ms MAC frame, which is the longest possible burst in the HIPERLAN / 2 system, the displacement of time at the end of the burst will be 2ms * 40ppm = 80ns. The phase factor for the highest used subcarrier that is at 8.125 MHz will then be \ omegaT_ {off} = 2 \ pi8. 125MHz * 80 ns - 4. 1 radian or 234 degrees. Such phase error will, therefore, cause the consistent demodulation is impossible.

A performance degradation will happen a lot before this depending on the link adaptation mode.

There are several prior art methods to compensate for the timing deviation. A first method offset of time deviation such is to tune suitably a reference oscillator in the receiver (using a timing deviation estimate) to make the Deviation from timing is negligible. A problem with This first method resides in the time it takes for the synthesizer RF adjust to a new value. Typically the RF synthesizer is implemented with a locked loop (PLL) phase phase), which needs establishment times that are large compared to the duration of the typical preamble. In addition, a voltage controlled reference will typically have more noise than phase (fluctuations in the frequency around an average) that a Oscillator reference that moves freely.

As mentioned before, in order to perform a consistent demodulation the receiver has to calculate a channel estimate, initially calculated in the preamble. From according to a second method of offsetting deviation from prior art time, the initial channel estimate is monitored and corrected with the help of pilots or methods Directed by decision during reception. This way, you can there is compensation for slow changes such as deviation from the timing

However, this approach requires a complex tracking mechanism, when tracking of each subcarrier separately, and susceptibility to dipoles that fade, when only a small number of pilots are used (HIPERLAN / 2 only has 4 pilots of the 52 subcarriers).

Deviation compensation techniques of the Timing are known in the sector.

EP 1049301 of the prior art, titled "Sampling clock correction in a multicarrier receiver ", is directed to an OFDM receiver structure that makes it possible to use an NCO (Numerically Controlled Oscillator) (Numerically Controlled Oscillator) to generate the clock of Sampling that is cheaper than VCO (Voltage Controlled Conventional Oscillator) (Tension Controlled Oscillator) conventionally used to achieve synchronization between transmitter and receiver clocks. The receiver uses the correction both the phase error of the sample clock and the error of frequency in the frequency domain. The estimation of cited errors use pilot subcarriers. This document forms the preamble of claim 1.

WO 98/39886 of the prior art entitled "Improvements in, or relating to synchronization", use frequency domain data to estimate the timing of OFDM symbol. The timing is quite rude and is intended that ensures symbol timing starts somewhere within the cyclic prefix. Using several starting points of Candidate symbol, a so-called objective function is evaluated. The candidate starting point that provides the lowest value of the Objective function is chosen as the symbol start point. It is recognized that a timing error causes a slope linear of the subcarrier arguments, the so-called function of argument. The method according to the previous document uses a estimation of the perturbation of the argument function as a objective function.

Summary of the Invention

A first object of the invention is to establish a radio receiver system that optimizes saving periods of energy where the receiver is not intended to receive data.

This object has been achieved through claim 1.

In other aspects of the invention, the system of radio receiver comprises a radio receiver, a clock of receiver sample, and a timing correction unit. He radio receiver receives several radio frequency carriers modulated and produces a baseband signal from them modulated according to a sample rate set by the Sample clock of the receiver. The correction unit of timing performs, in the frequency domain, a compensation of the timing deviation between a clock of transmitter sample and receiver sample clock for improve, for example, the demodulation carried out by a unit of demodulation.

In different embodiments, the deviation from the timing can be carried out together with systems that they use differing techniques, such as an estimation technique channel or signal compensation technique. With each technique realization, estimation techniques of the frequency deviation that differ, such as (for example) estimation of frequency deviation directed by preamble, estimate of the frequency deviation directed by decision, or estimate of the frequency deviation helped by a pilot.

In exemplary embodiments, the receiver includes a demodulation section comprising the demodulator unit  and the timing correction unit. The receptor it also includes a deviation estimation unit of frequency; a frequency correction unit; a unit of fast Fourier Transform (FTT) (fast Fourier transform); and one channel estimation unit.

In channel estimation embodiments of example, the unit of frequency deviation estimation provides an estimate of the frequency deviation. The frequency correction unit receives the baseband signal modulated and estimated frequency deviation, and provides a frequency modulated baseband signal corrected The fast Fourier transform (FFT) unit Fast Fourier) receives the modulated baseband signal from frequency corrected and provides, for each subcarrier of the multi-carrier, a signal modulated in the frequency domain. The channel estimation unit uses the baseband signal frequency modulated corrected to generate an estimate of channel in the frequency domain. The channel estimate in the frequency domain is applied to the correction unit of timing The timing correction unit generates a channel frequency domain frequency estimation corrected that is applied to the demodulator. The demodulator uses the modulated signal of the frequency domain and the estimation of channel in the time frequency domain corrected for generate a demodulated signal.

In the channel estimation embodiments, the timing correction unit estimates a deviation value of timing and compensates for the deviation value of the timing in the frequency domain by applying a factor of appropriate phase to a subcarrier to update the estimate of channel in the frequency domain and therefore provide a channel estimation in the time frequency domain corrected.

In exemplary compensated signal embodiments, the offset of the timing deviation is brought to out based on the estimate of the frequency deviation and in a modulated signal of the frequency domain. The unit of frequency deviation estimate provides an estimate of frequency deviation The correction unit of frequency receives the modulated baseband signal and the estimate  of frequency deviation, and provides a band signal Modulated base frequency corrected. The fast Fourier unit transform (FFT) (fast Fourier transform) receives the signal from frequency modulated baseband corrected and, for each subcarrier, provides a modulated signal in the domain of the frequency that is applied to the demodulator. The unit of estimation of channel uses the frequency modulated baseband signal corrected to generate a channel estimate in the domain of the frequency. The timing correction unit receives the estimate of frequency deviation and modulated signal of the frequency domain, and generates a modulated signal in the domain of the corrected time frequency that applies to demodulator The demodulator uses the modulated signal from the domain of the corrected time frequency and the channel estimate for generate a demodulated signal.

In the embodiments, the compensated signal of example, the timing correction unit estimates a value of the timing deviation and compensates for the value of the timing deviation in the frequency domain applying an appropriate phase factor to a subcarrier to update the frequency domain modulated signal, thereby providing a modulated signal in the domain of the Corrected time frequency.

       \ newpage
    

In an example implementation context, the different carriers of modulated radio frequencies have been modulated using Orthogonal Frequency Division Multiplexing (OFDM) (Orthogonal Frequency Division Multiplexing). Agree with an optional mode of the invention, updating the channel estimate for offset deviation compensation timing occurs every Mi symbol (a value being selected for M that depends on a link adaptation mode particular).

In another of its aspects, the present invention refers to a mobile station that includes the radio receiver and the receiver's sample clock, as well as an estimation unit of frequency deviation. The deviation estimation unit frequency receives the modulated baseband signal from the radio receiver and provides an estimate of the deviation of the frequency. The mobile station also comprises a controller which uses the frequency deviation estimate to determine a corresponding sample clock deviation (where the corresponding sample clock deviation is estimated a deviation between the sample clock of the receiver and the clock of transmitter sample). In one embodiment, the controller is a sleeping mode controller that also determines a deviation of the timing during a sleep period of a duration default In addition, the controller in sleeping mode determines one or more of the following: a time until the receiver is going to sleep, and a while until the receiver will look for a start of a plot.

Brief description of the drawings

The above and other objects, characteristics and advantages of the invention will be apparent from the description. more detailed particular of preferred embodiments as illustrated in the accompanying drawings in which characters referenced refer to the same parts in the different views. The drawings are not necessarily to scale, putting in instead place emphasis on illustrating the principles of the invention.

Fig. 1 is a diagrammatic view of a frame structure and an example burst transmitted over the air interface of HIPERLAN / 2.

Fig. 2 is a more detailed diagram view. of portions of an example burst of Fig. 1.

Fig. 3 is a schematic view of a Sample radio receiver receiving modulated OFDM signals from a transmitter, and that performs a compensation of the Deviation from the timing according to this invention.

Fig. 4 is a schematic view of a demodulation section of a radio receiver that uses the channel estimation along with offset deviation compensation timing, and that uses an estimate of the deviation from the frequency directed by preamble.

Fig. 5 is a schematic view of a demodulation section of a radio receiver that uses the channel estimation along with offset deviation compensation timing, and that uses an estimate of the deviation from the frequency directed by preamble.

Fig. 6 is a schematic view of a demodulation section of a radio receiver that uses the channel estimation along with offset deviation compensation timing, and that uses an estimate of the deviation from the frequency directed by preamble.

Fig. 7 is a schematic view of a demodulation section of a radio receiver that uses the channel estimation along with offset deviation compensation timing, and that uses an estimate of the deviation from the frequency directed by preamble.

Fig. 8 is a flow chart showing basic example actions carried out by a directed unit by decision according to an embodiment of the invention.

Fig. 9 is a schematic view of a Sample radio receiver receiving modulated OFDM signals from a transmitter, and that performs a compensation of the Deviation from the timing after a sleep period of according to an example mode of the present invention.

Fig. 10 is a schematic view of portions of the radio receiver of Fig. 9, which particularly shows elements that facilitate compensation for deviation from the Timing after a period of sleep.

Fig. 11 is a flow chart showing certain representative, basic, example steps carried out by a sleep phase controller used in the receiver of radius of Fig. 9 and Fig. 10.

Detailed description of the invention

In the following description, for the purpose of explanation and not limitation, specific details are established such as architectures, interfaces, techniques, etc. private individuals in order to provide a complete knowledge of this invention. However, it will be evident to experts in the technique that the present invention can be put into practice in other embodiments that are separated from these details specific. In other examples, detailed descriptions of well-known devices, circuits and methods are omitted with in order not to obscure the description of the present invention with An unnecessary detail. In addition, individual function blocks are show in some of the figures. Those skilled in the art they will appreciate that the functions can be implemented using individual hardware circuits, using software that works together with a properly programmed digital microprocessor or a general purpose computer, using an application specific integrated circuit (ASIC) (application-specific integrated circuit), and / or using one or more digital signal processors (DSPs) digital signal).

Fig. 3 shows a radio receiver 30 of example that receives, on air interface 32, a plurality of carriers of modulated radio frequencies 34 that are transmitted from antenna 36 of transmitter 38. Radio receiver 30 of example can be called or take the form of mobile stations such as user equipment units (UEs, phones mobile phones (cell phones) and laptops with mobile termination, and thus can be, for example, portable, pocket, manual, computer included, or mobile devices mounted on a car that communicate voice and / or data with the network of radio access In an example implementation context, the different carriers of modulated radio frequencies have been modulated using Orthogonal Frequency Division Multiplexing 34 (OFDM) (Multiplexing by Orthogonal Frequency Division). With illustration views, radio frequency carriers modulated 34 are transmitted according to the standard HIPERLAN / 2, although the present invention can be used with any OFDM system.

As also shown in Fig. 3, the radio receiver 30 includes an antenna 40 that is connected to a radio receiver section 41. The radio receiver section 41 receives the modulated radio frequency carriers 34 and applies an analog complex modulated baseband signal on the line 42 to an analog to digital converter (ADC) analog to digital) 43. The ADC 43 in turn applies a signal from digital complex modulated baseband (discrete time) in the line 44 (for example, complex baseband OFDM symbols) to an initial timing synchronization section 46. The synchronization section of initial timing 46 performs an estimate of the start of each burst of a frame (see Fig.  one). The digital complex modulated baseband signal is applied to the demodulation section 50 which, for each subcarrier, provides a demodulated but interleaved signal. The exit of the demodulation section 50 is applied in sequence to the section de-interleaving 54, to the section of decoding FEC 56, and to the section of de-randomization 58.

ADC 43 performs signal sampling received The sampling carried out by the ADC 43 takes place at a sampling rate governed by the sample clock of the receiver 60. At this end, Fig. 3 shows the receiver of radius 30 also comprising the sample clock of the receiver 60 which is connected to ADC 43. Sampling of the signal in the receiver 30 is thus carried out at the clock speed of Receiver sample 60 while signal processing it was carried out by transmitter 38 according to a clock of transmitter sample 66.

A deviation from the potential timing that the present invention compensates takes place when the signal is converted by ADC 43 from continuous domain over time, analog to discrete time domain, digital, that is, when The signal is sampled. The timing deviation is inherent in the signal once it is sampled. In a DSP implementation samples can be stored in a memory. Once the entire burst has been stored, the section of demodulator 50 treats it, for example, at a much higher speed than the speed of the sample.

As explained in more detail here in the context of different example embodiments, the section of demodulation 50 includes a demodulation unit that leads to perform a demodulation of a modulated subcarrier signal obtained from the modulated baseband signal. In view of a potential timing deviation between the sample clock of receiver 60 and the sample clock of transmitter 66, the section Demodulation 50 includes a correction unit timing or timing correction function 100 that carries out, in the frequency domain, a compensation of the timing deviation (to compensate for the deviation of the timing between the transmitter's sample clock and the clock receiver sample), thereby improving demodulation in the radio receiver 30.

In different embodiments, the compensation of timing deviation can be carried out from According to different techniques. Representative techniques, not Limitations, for example, described below include the offset of timing deviation based on channel estimation and offset deviation from the applied signal timing (i.e. a signal technique compensated). In addition, with each embodiment, they can be used different techniques for estimating frequency deviation, such as (for example) estimate of the deviation of the frequency directed by preamble, estimation of the deviation of the frequency directed by decision, or estimate of the deviation of the frequency helped by a pilot.

For each such example embodiment, the receiver 30 includes a demodulation section 50, section that comprises the demodulator unit and the correction unit of timing As used here, the numbers of reference without suffix 50 and 100 represent a section of 50 generic demodulation and a correction unit 100 generic timing, respectively. Similarly numbered elements of the different embodiments have essentially identical functions that respect aspects of timing correction of the present invention.

Fig. 4 illustrates basic example components of a radio receiver 30 that has a demodulation section 50 (4) that uses the channel estimate and the estimate of frequency deviation directed by preamble. In Fig. 4, the digital complex modulated baseband signal R_ {G} [1] obtained on line 48 from the section of initial timing synchronization 46 is applied both in frequency estimation unit directed by preamble 102 as in the frequency correction unit 104.

As those skilled in the art understand, a frequency deviation takes place due to the limited accuracy of frequency generation in the transmitter and / or receiver radio Such frequency deviation destroys the orthogonality of OFDM subcarriers, and according to this should be estimated and removed from the received signal in order not to degrade performance

According to this, the unit of estimation of frequency deviation directed by preamble 102 generates a estimate of the deviation of the frequency that is extracted in the line 106 towards the frequency correction unit 104 and the timing correction unit 100. The estimation unit frequency-driven deviation by preamble 102 uses the preamble (see preamble 22 of Fig. 1 and Fig. 2), and you need to know when the preamble begins, in order to generate the Frequency deviation estimate. In view of this, it is assumed that the signal R_ {G} [1] on line 48 has been aligned (for example, by the synchronization section of the initial timing 46) so that the preamble begins at the first sample that enters the unit of estimation of the frequency deviation directed by preamble 102. The unit of frequency deviation estimation directed by preamble 102 can generate the estimate of the frequency deviation using any of several techniques, which include self-correlation (which exploits the periodicity of the preamble) and cross correlation (in which, for example, a received preamble is cross correlated with a stored reference). In essence, in the preamble the parts 22_ {2} and 22_ {3} are a repetition of the same sequence, and correlating them, for example, by carrying out auto-correlation, frequency deviation It can be estimated. Frequency correction unit 104 uses the estimate of the frequency deviation obtained from the unit of frequency deviation estimation directed by preamble 102 to generate a modulated baseband signal of corrected frequency shown on lines 108 of Fig. 4.

The demodulation section 50 (4) It also includes a Fast Fourier Transform Unit (FFT) 110; one channel estimation unit. 112 the correction unit timing 100 (4); and a demodulation unit 114 (4). Fast Fourier Transform (FFT) 110 receives the frequency band modulated baseband signal applied to it on line 108, and provides, for each subcarrier, a signal modulated in the frequency domain shown by the line 116. The channel estimation unit 112 uses the band signal of frequency modulated base applied to it on the line 108 to generate a channel estimate in the domain of the frequency. The channel estimate in the frequency domain, shown on line 118 and the notation H_ {m} [0] of the Fig. 4, is applied to the timing correction unit 100 (4). The timing correction unit 100 generates a channel estimate in the time frequency domain corrected which is shown by the notation H_ {m [} k] and is applied on line 120 to demodulator 114 (4). The demodulator 114 (4) use the frequency domain modulated signal (obtained from the Fast Fourier transform (FFT) 110) and the channel estimate in the time frequency domain corrected (obtained from the correction unit of timing 100 (4)) to generate a demodulated signal. The demodulated signal is extracted on line 122 towards the section of interleaving 54 (see Fig. 3).

In the embodiment of the channel estimate represented by Fig. 4, the timing correction unit 100 (4) estimates a value of the timing deviation t_ {0} and compensates the value of the timing offset in the frequency domain by applying an appropriate phase factor \ varphi_ {m, k} to a subcarrier to update the channel estimate in the frequency domain and thereby provide a channel estimate in the corrected time frequency domain H_ {m} [k ] on line 120 to a demodulation unit 114 (4). An explanation is provided that provides a basis for such operation of the timing correction unit 100.

As mentioned above, a time offset T_ {off} of a function f (t) in the time domain provides a linear phase factor of the corresponding Fourier transform as indicated in Equation 2.

       \ vskip1.000000 \ baselineskip
    

2

       \ vskip1.000000 \ baselineskip
    

This is compared to the Fourier transform F (T) of the original function f (t) according to Equation 3.

       \ vskip1.000000 \ baselineskip
    

3

       \ newpage
    

Since there is a deviation from the timing the offset, T_ {off} , grows linearly with time according to Equation 4.

4

In Equation 4, t_ {0} is the deviation of the corresponding sample clock, -40 ppm < t_ {0} <40 ppm.

Assuming that T_ {off} changes very slowly, Equation 3 still remains roughly. The phase factor can then be written as shown in Equation 5.

5

The phase factor can also be written in terms of symbol number k and subcarrier index m according to Equation 6.

6

In Equation 6, m is the index of the subcarrier for the subcarriers used, - 26 ≤ m ≤ 26; k is the time index measured in OFDM symbols, with k = 1 being the first data symbol, that is, the first symbol after the preamble, k = 1, 2, 3 ...; T_ {s} is the time of the OFDM symbol, T_ {s} = 4 \ mus (80 samples); T is the sample time, T = 50 ns; and T_ {init} is the time between the reference time t = 0 and the first data symbol.

As shown by Equation 7, the estimation of the timing deviation t_ {0} can be derived from the estimation of the frequency deviation since the radio frequency and the sample clock are derived from it reference source, according to the HIPERLAN / 2 standard. The estimate of the frequency deviation is taken in any way at the beginning of the burst to perform frequency synchronization, so that it does not require extraordinary calculations.

7

In Equation 7, f_ {off} is the estimate of the absolute frequency deviation in Hz, and f_ {c} is the carrier frequency in Hz.

Once the timing deviation is estimated, can be compensated in the frequency domain applying the correct phase factor to each subcarrier. This is does after the fast Fourier transform (FFT) and before the demodulation

The embodiment of Fig. 4 reflects the method of update the channel estimate according to Equation 8.

8

In Equation 8, H_ {m} [k] is the channel estimate in the domain of the corrected complex frequency of the subcarrier m ava of the symbol k of OFDM that carries data, and H_ {m } [0] is the channel estimate in the domain of the initial complex frequency of the subcarrier m ava .

From the foregoing it is understood that, in the embodiment of Fig. 4, the timing correction unit 100 (4) updates the channel estimate using the ratio H_ {m} [k ] = exp ( j \ cdot \ phi_ { m, k} ) \ cdotHm [0], in which: H_ {m} [k ] is the corrected time deviation channel estimate for a time index measured in the data symbols k; H_ {m} [0] is the initial channel estimate; \ varphi_ {m, k} is the phase factor; and in which m is a subcarrier index for the subcarriers used. The factor phase \ varphi_ {m, k} is in turn defined by 9 where: m is the subcarrier index; k is the time index measured in data symbols; T_ {s} is a time symbol; T is a sample time; T_ {init} is a time between a reference time and a first data symbol; and t_ {0} is the value of the timing deviation. The value of the timing deviation t_ {0} is derived from 10 where f_ {off} is an estimate of the deviation of the absolute frequency in Hz, and f_ {c} is a carrier frequency in Hz.

The update carried out by the timing correction unit 100 (4) to compensate for the deviation of the timing only needs to be performed each symbol M ^ in order to save calculations since the timing deviation is quite a process slow. At this end, Fig. 4 shows an optional variation of the inclusion in Fig. 4 of a controller 130. The controller 130, shown by the dashed lines reflects its optional inclusion, coordinates the action of the timing correction unit 100 (4) so that the update of the channel estimate for compensation of the timing deviation takes place every symbol M ^. Fig. 4 particularly shows a signal M that goes from controller 130 to timing correction unit 100 (4), advising on how many symbols are going to use the same channel estimate. The update period M may, for example, depend on the link adaptation mode, since the more sensitive, higher modulation schemes will require a faster update.

Thus, controller 30 applies H_ {m} [1] for the first symbols M, H_ {m} [M + 1] for the following symbols M and so on. Mathematically, this update has the effect shown in Equation 9, in which n \ in N (natural numbers) and p is an element of [ 1, M ]. The activation signal M ava is thus applied to the timing correction unit 100, telling it to update the channel estimate.

eleven

Fig. 5 illustrates basic example components of a radio receiver 30 that has a demodulation section 50 (5) that applies the compensation to the signal and uses the estimate of frequency deviation directed by preamble. In the example compensated signal embodiment, the offset of timing deviation is carried out based on the frequency estimate and is applied to the signal frequency domain modulation. Where appropriate, numbers reference components of the embodiment of Fig. 5 which are basically similar to the corresponding components of the Embodiments of Fig. 4 are the same as in Fig. 4

As in the embodiment of Fig. 4, in Fig. 5 the unit of frequency deviation estimation directed by preamble 102 receives the modulated baseband signal R_ {G} [I] and provides (on line 106) an estimate of frequency deviation The correction unit of frequency 104 receives the modulated baseband signal and the estimate of the frequency deviation and provides (in the line 108) a frequency modulated baseband signal corrected The Fast Fourier Transform (FFT) 110 receives the digital frequency complex modulated baseband signal corrected and, for each subcarrier, provides a modulated signal (separated) from the frequency domain R_ {FFT, m} [k] which is applied to the timing correction unit 100 (5).

The channel estimation unit 112 of Fig. 5 uses the frequency band modulated baseband signal to generate a channel estimate in the frequency domain which is applied on line 140 to the demodulation unit 114 (5). In the embodiment of Fig. 5, the unit of timing correction 100 (5) receives the modulated signal of the frequency domain R_ {FFT, m} [k] and the estimate of the frequency deviation (on line 106). The unit of timing correction 100 (5) generates a modulated signal from the corrected time frequency domain R_ {TD, m} [k] which is applied to demodulator 114 (5). The unit of demodulation 114 (5) uses the modulated signal from the domain of the Corrected time frequency R_ {TD, m} [k] and estimate channel to generate a demodulated signal that is applied in the line 122 to de-interleaving section 54.

Thus, in the example signal compensation embodiment of Fig. 5, the timing correction unit 100 (5) estimates a value of the timing deviation for and compensates the value of the timing deviation in the domain of the frequency by applying an appropriate phase factor var m, k to the subcarrier in order to update the modulated signal of the frequency domain, thereby providing a modulated signal of the corrected time frequency domain R_ { TD, m} [k] to the demodulation unit 114 (5).

The method implemented by the realization of the Fig. 5 corrects the signal in this way instead of estimating channel, using Equation 10.

12

In Equation 10, R_ {TD, m} [k] is a signal of the compensated complex frequency domain for the timing deviation of the subcarrier m ava of the kd OFDM symbol that carries data, and R_ {FFT, m} [k] is the signal of the complex frequency domain just after the FFT of the subcarrier m ava of the k avo OFDM symbol that carries data. As in the embodiment of Fig. 4, the phase factor is understood with reference to Equation 5 and Equation 6, and the estimation of the timing deviation t_ {0} is understood, for example, with reference to the Equation 7.

Instead of using the frequency deviation estimate, the timing deviation t_ {0} can also be estimated by a decision driven method and / or assisted by a pilot. For example, Fig. 6 illustrates basic example components of a radio receiver 30 having a demodulation section 50 (6) that utilizes channel estimation and estimation of the frequency deviation directed by decision. Where appropriate, reference numbers of components of the embodiment of Fig. 6 that are basically similar to the corresponding components of the embodiments of Fig. 4 are the same as in Fig. 4 Notably in Fig. 6 The demodulation section 50 (6) includes a decision-driven unit, also referred to herein as the decision-driven frequency deviation estimation unit 150 (6).

As another example, Fig. 7 illustrates components sample basics of a radio receiver 30 that has a section of demodulation 50 (7) that applies the compensation to the signal and that uses frequency deviation estimation directed by decision. Again, where appropriate, reference numbers of components of the embodiment of Fig. which are basically similar to the corresponding components of the embodiments of Fig. 5 are the same as in Fig. 5 The section of demodulation 50 (7) of the embodiment of Fig. 7 includes a decision-driven unit (deviation estimation unit frequency directed by decision 150 (7)).

The operating principles and details of the unit of frequency deviation estimation directed by decision 150 (6) of Fig. 6 and the unit of estimation of the frequency deviation directed by decision 150 (7) of the Fig. 7 are understood with respect to a unit directed by representative example decision 150 illustrated in Fig. 8. In the embodiments of Fig. 6 and Fig. 7, for purposes of the present invention, the decision driven unit 150 (6) and 150 (7) each have three inputs and one output. The three entries are shown on the lines that have numbers of base reference 152, 154, and 156, while the output is found on a line that has the base reference number 158. For example, decision-driven unit 150 (6) of Fig. 6 has a first entry on line 152 (6) of the demodulation unit output 114 (6); A second entry on line 156 (6) of the output of the unit timing correction 100 (6); a third entry in line 154 (6) of the Fourier transform output Fast (FFT) 110 (6); and an output applied on the line 158 (6) to the timing correction unit 100 (6). The decision-driven unit 150 (7) of the Fig. 7 has similar connections, with the exception of its second input line 156 (7) that goes from an output of the unit of channel estimation 112 (7) and its third input line 154 (7) that goes from the timing correction unit 100 (7).

It is recalled that the timing correction unit 100 (4) of Fig. 4 and the timing correction unit 100 (5) of Fig. 5 use an estimate of frequency deviation directed by preamble f_ {off} with the in order to determine the timing deviation t_ {0} . In the decision-directed embodiments of Fig. 6 and Fig. 7, on the other hand, the decision-directed unit 150 (6) and the decision-directed unit 150 (7), respectively, provide the correction unit of timing 100 (6) and timing correction unit 100 (7) estimating the frequency deviation f_ {off} as described below.

The decision-driven unit 150 as shown in Fig. 8 comprises a correlation unit 162 that receives an input from the demodulation unit 114 on line 152. A first input terminal of multiplier 164 receives a first input on the line 156 (from the timing correction unit 100 of the embodiment of Fig. 6 and from the channel estimation unit 112 of the embodiment of Fig. 7) and a second input from the output of the correlation unit 162. The output of multiplier 164, shown as Bm [k] in Fig. 8, is applied to the first input of the phase discrimination unit 166. A second input to the phase discrimination unit 166 is received on the line of input 154 from the FFT unit 110 of the embodiment of Fig. 6 and from the timing correction unit 100 of the embodiment of Fig. 7.

The output of the phase discrimination unit 166 (shown as \ varphi_ {est} [k] in Fig. 8) is applied to the first order loop filter 168. An output of the first order loop filter 168 is connected with integrator 170. The output of integrator 170, shown as \ varphi_ {A} [k] in Fig. 8, is applied to delay unit 172. Output 172, shown as \ varphi_ {A} [k -1] in Fig. 8, is applied to the carrier phase deviation calculation unit per sample 174. The output of the carrier phase deviation calculation unit per sample 174, shown as \ varphi_ {s} [k-1] in Fig. 8, is applied to the frequency deviation estimation unit 176. The output of the frequency deviation estimation unit 176, that is, f_ {off} [k-1] , it is extracted on line 158 towards timing correction unit 100.

The decision driven unit 150 receives a signal comprising a series of data symbols, and calculates an estimated phase deviation for each data symbol as a function of the respective data symbols. Furthermore, as a function of the estimated phase deviation of a data symbol and the estimated phase deviation of a data symbol that precedes the last data symbol, a predicted phase deviation for the data symbol in question is calculated. From the predicted phase deviation the decision driven unit 150, and particularly the frequency deviation estimation unit 176, calculates the estimate of the frequency deviation f_ {off} . From the estimation of the frequency deviation f_ {off} , the timing correction unit 100 can calculate the timing deviation t_ {0} , for example, using Equation 7.

Fig. 2 illustrates the structure of a series of Received samples that include the phase reference points. OFDM symbols 22_ {2} and 22_ {3} of Preamble-C are used for a channel estimate and an estimate of the phase reference (both symbols being OFDM 22_ {2} and 223 used for channel estimation and the phase of reference). Before the actual channel estimation, the two symbols of the preambles-C are added to obtain a major noise suppression As a result of this average process, the phase reference point of the R_ {CE} channel estimate it is located in the middle of the OFDM symbols 22_ {2} and 22_ {3} in the domain of time. The actual data series of symbols from data (ie OFDM symbols) follows the preamble-C. In what follows the actual data series It is also called burst, where each burst comprises several symbols OFDM preceded by a preamble-C.

Among its other functions, the directed unit by decision 150 estimate the phase deviation of the first symbol of OFDM data S1. Assuming a constant frequency deviation, the estimated phase deviation \ varphi_ {est} [1] corresponds to the phase deviation in the middle of R_ {S1} of the OFDM S1 symbol in the time domain. The difference between the reference points R_ {CE} and R_ {S1} in the domain of time is denoted by y_ {1} in Fig. 2. Based on the phase deviation \ varphi_ {est} [1], the follower of frequency calculates the phase of de-rotation \ varphi_ {corr, 0} [2]. The phase of des-rotation \ varphi_ {corr, 0} [2] is corresponds to a phase deviation \ varphi_ {A} [1] to principle S_ {S2} of the second OFDM symbol S2. This deviation phase \ varphi_ {A} [1] is a predicted phase deviation for the second OFDM S2 symbol. In the domain of time, the differences between the reference point R_ {ce} and the start S_ {k} of OFDM symbols that constitute a point of phase reference for each OFDM symbol S_ {k} is called x_ {k}. The parameters x_ {k} and y_ {1} are used to determine the optimal coefficients for the phase locked loop of the frequency follower In addition, it is observed that the so-called deviation predicted phase \ varphi_ {A} [k-1] represents the phase increment from the reference point of R_ {ce} phase of the preamble-C at the start S_ {Sk} of the OFDM symbol k-avo S_ {k}. Deviation from the phase correction \ varphi_ {corr, 1} [k] represents the phase increment from the phase reference point R_ {ce} with regarding the 1-a sample of the k-avo OFDM symbol S_ {k}.

As used here, the index "k" indicates the number of symbols in a burst, while the index "C" is use to distinguish signals. The index "1" represents the number of samples derived for each OFDM symbol.

The demodulation unit of the subcarrier output 114 is remodulated by correlation unit 162 to obtain the remodulated symbols A_ {m} [k]. The unit of correlation 162 performs output remodulation u [k] by means of a correlation carried out in accordance with the standard of HIPERLAN / 2. The remodulated symbols A_ {m} [k] are multiplied by multiplier 164 with the entry on line 156 (for example, channel estimation H_ {m} [k] of each subcarrier to carry out the Fig. 8). For the embodiment of Fig. 8, therefore, the multiplication carried out by multiplier 164 is like the shown in Equation 11 to obtain a symbol estimate weighted B_ {m} [k] on each subcarrier. In Equation 11, "m" designates the number of subcarriers.

13

Phase discrimination between samples of received signal R_ {m} [k] and the symbol estimate weighted B_ {m} [k] produces the phase estimate \ varphi_ {est} [k] extracted by the discrimination unit of phase 166. The phase estimate \ varphi_ {est} [k] is also calculated by the filter F (z), which comprises a loop First-order filter 168 which has a transfer function described by Equation 12. Equation 13 describes F (z). Equation 14 defines a calculation of the filter output F (z), where \ varphi_ {A} [0] and \ varphi_ {A} [k-1] are zeroed with initialization purposes

14

       \ vskip1.000000 \ baselineskip
    

fifteen

       \ vskip1.000000 \ baselineskip
    

16

The exit \ varphi_ {A} [k-1] representing the phase increment from the phase reference point R_ of the preamble-C to the start S_ {Sk} of the symbol of OFDM k-avo Sk is sent to delay unit 172. Thus, an H_ {2} filter [z] is obtained which is of the type of second order, in which the filter transfer function H_ {2} [z] as open loop is described by Equation fifteen.

       \ vskip1.000000 \ baselineskip
    

17

The exit \ varphi_ {A} [k-1] of filter F (z) corresponds to the phase that would exist in the OFDM symbol k-avo Sk if no tracking of frequency for a received OFDM signal that has a low noise. For the sake of simplicity, in what follows, a constant frequency deviation. However, the following bypass is still valid when frequency deviation constant corresponds to the average frequency deviation in a certain time, which is typically an OFDM symbol, in case of a frequency deviation that varies for example due to noise phase As a result, phase deviation \ varphi_ {s} [k] per sample of signal extracted by means 24 is given by Equation 16.

18

Since the phase \ varphi_ {A} [k] represents a predicted phase deviation an OFDM symbol respective, the sample phase deviation \ varphi_ {s} [k] is also a predicted phase deviation since it is calculated as a function of the phase \ varphi_ {A} [k]. In order to take into account the underlying prediction property, phase deviation of sample \ varphi_ {s} [k] is obtained by a division of the phase deviation \ varphi_ {A} [k] by the number x_ {k + 1} of samples between the reference point R_ {CE} and the start S_ {Sk + 1} in the time domain of the OFDM symbol S_ {k + 1} comprising the samples in question.

Sample phase deviation \ varphi_ {s} [k] predicted is applied to the unit of frequency deviation estimate 176. The unit of estimation of frequency deviation 176 performs a calculation of the frequency deviation using Equation 17. In Equation 17, T is the sample time.

19

As previously indicated, the timing correction unit 100, using the frequency deviation f_ {off} provided by the decision driven unit 150, can calculate the timing deviation t_ {0} according to Equation 7. Thus, the decision driven unit 150 (6) and the decision driven unit 150 (7) provide alternative ways of providing the timing deviation t_ {0} , without having a delay in estimating the deviation of the frequency directed by preamble.

The above describes the case if the frequency correction unit 104 is not applied. For the case in which the frequency correction unit 104 is applied, the frequency deviation estimated by the previation-driven frequency deviation estimation unit 102 must be added to the frequency deviation calculated by the decision-directed unit. 150. In this case, the decision-driven frequency estimation unit 150 is an additional means for refining the unit of frequency-driven estimation of preamble 102. As indicated in Fig. 6 and Fig. 7, the unit The timing correction unit has connections to the frequency deviation estimation unit directed by preamble 102 and to the decision directed frequency deviation estimation unit 150, which means that the timing correction unit 100 obtains two values for the frequency deviation, which must be added. Therefore, Equation 7 must be modified for this case in which two values are obtained for f_ {off} , a value ( f_ {offpreambule} ) from the frequency deviation estimation unit directed by preamble 102 and another value ( f_ {offdecision} ) from the frequency deviation estimation unit directed by decision 150, and taken into account when calculating t_ {0} according to the following: t_ {0} = ( f_ {offpreambule} + ( f_ {offdecision}) / f_ {c} ].

As another alternative approach, it you can use the frequency deviation estimate Helped by pilot. In this context "pilot aided" wants say that at least some pilots and symbols are transmitted, which They are known on the receiver side. For example, some values A_ {m} [k] represent estimates of the symbols of the pilot. In correlation unit 162 of Fig. 8 you are Symbol estimates are replaced by pilot symbols known. This substitution avoids decision errors, which could lead to a false estimate of the deviation of the frequency. The drawback of an estimate of deviation from the Pilot-assisted frequency lies in the redundancy of the Pilot symbols, which reduces the overall data rate of the system.

Thus, advantageously embodiments of the present invention can be implemented using a frequency of reference that runs freely that does not have to be adjusted during reception and is characterized by low noise phase. Since the frequency deviation has to be estimated in any way, the present invention advantageously re-use frequency deviation since it is identical to the deviation of the sample clock. Also, with the present invention the complexity of the compensation of the Deviation from the required timing is kept to a minimum and It is suitable for implementation purposes.

The solution adopted in the present invention provides compensation for timing deviation while receiving a burst of PHY. This solution avoids changing the frequency oscillator reference, and can be implemented with digital signal treatment, but also with complexity Very small and high robustness.

Fig. 9 shows a radio receiver 30 that includes a sleep phase controller 200 that provides compensation for the timing deviation after a sleep period. The selected elements of the radio receiver 30 that facilitate compensation of the timing deviation after a sleep period are shown in more detail in Fig. 10. The radio receiver 30 of the embodiment of Fig. 9 and the Fig. 10 compensates for the timing deviation in a situation in which the radio receiver 30 goes to sleep for a certain number of frames, then wakes up to receive a frame, and then either sleeps again or continues to receive frames. It should be understood that the sleep phase controller 200 of Fig. 9 and Fig. 10 can be used in conjunction with many shapes and variation of the demodulation section, which include the specific demodulation sections 50 previously described. For the purpose of simplification, only the frequency offset estimation unit directed by preamble 102 and the frequency correction unit 104 of the demodulation section 50 are shown in Fig. 10, it being understood that other elements are also included but not They are necessary for the explanation of the sleep phase controller 200 per se .

Several problems have been found in a sleep period scenario. A first problem such is that the timing deviation during sleep has to be Dear to know when the receiver has to wake up. A second problem such is that the accuracy of the estimate has to be known in order to properly design a window of search, that is the window during which a plot start search. These problems are not relevant. only for OFDM systems, but for all systems with mode of dream.

Fig. 11 shows certain example steps, basic carried out by the sleep phase controller 200 of according to a mode of the dream mode aspect of the present invention. As explained below, the phase controller of sleep 200 determines the timing deviation T_ {off} in seconds during a sleep period, determine a time up to which the receiver (for example, mobile station) can sleep, and determines a start of frame search up to time (a time for which the search for the beginning of a frame is going to continue).

Step 11-1 of the procedure of Fig. 11 carried out by the sleep phase controller 200 shows the sleep phase controller 200 receiving the estimate of the frequency deviation f_ {off} from an estimation unit of appropriate frequency deviation (for example, the frequency deviation estimation unit directed by preamble 102 of Fig. 10). After obtaining the estimate of the frequency deviation f_ {off} in step 11-1, in step 11-2 the sleep phase controller 200 derives a deviation from the corresponding clock sample estimated t_ {0} using Equation 7, previously provided.

As step 11-3, the sleep phase controller 200 uses the deviation of the corresponding corresponding sample clock t_ {0} to determine a deviation of the timing during sleep. In particular, the timing deviation T_ {off} in seconds during a sleep period of T_ {sleep} is calculated using Equation 18.

       \ vskip1.000000 \ baselineskip
    

twenty

In order to establish an appropriate search window, the size of the average search window must be determined. This implies determining the accuracy of the estimate t_ {0} , for example by knowing the estimation algorithm. If it is assumed that t_ {0res} is the maximum error of the estimate t_ {0} expressed in relation to the frequency of the sample clock, then the estimate of the deviation of is t_ {0} +/- t_ {0res} per example 40 ppm +/- 2 ppm. From this information as step 11-4 the sleep phase controller 200 can determine half a search window using Equation 19.

twenty-one

Then, counting from the time t = 0 when the sleep mode enters, as step 11-5 the sleep phase controller 200 determines that the mobile terminal must sleep until the time t expressed by Equation 20, and searches for the start of the plot until the time of Equation 21.

22

       \ vskip1.000000 \ baselineskip
    

2. 3

Thus, the sleep phase controller 200 determines the deviation of the Toff timing in seconds during  a period of sleep determines a time until which the recipient (for example, mobile station) can sleep, and determines a frame start search up to time (a time for which the search for the beginning of a frame will continue).

Although HIPERLAN / 2 is used here as a system of For example, the present invention in its different aspects may be used in any OFDM system.

The components that constitute the section of demodulation 50 in the different embodiments described previously referred to as "units" for ease reference. Alternatively, these units can be conceptualized in terms of functions. The units / functions of the demodulation section 50 can be implemented in various ways, which include using individual hardware circuits, using software that works together with a digital microprocessor Properly programmed or a general purpose computer, use an application specific integrated circuit (ASIC), and / or use one or several digital signal processors (DSPs). In addition, more than one of the units and functions as shown can be performed in any given hardware circuit or be made by means of the execution of a processor or similar. In addition, in this sense the timing correction unit can be performed by A controller or similar.

Claims (21)

1. A radio receiver apparatus that understands: a radio receiver (41) that receives different carriers of modulated radio frequencies and produces from there a modulated baseband signal, having been the different carriers of modulated radio frequencies transmitted by a radio transmitter (38) that operates according to a sample clock transmitter (66) in which the radio frequency and the clock of Sample are derived from the same source; a receiver sample clock (60) that is used to sample the modulated baseband signal; a deviation estimation unit of frequency (102, 150) that provides an estimate of the frequency deviation, t_ {0}; a timing correction unit (100), which performs in the frequency domain a compensation of the deviation of the timing between the sample clock of the transmitter and receiver sample clock. characterized in that the apparatus also comprises a sleeping mode controller (200) adapted so that get (11-1) an estimate of frequency deviation, f_ {off}; drift (11-2) a deviation from sample clock corresponding to the estimate, t_ {0}, being a deviation between the sample clock of the receiver and the clock of Transmitter sample of the following relationship: 24 where f_ {off} is an estimate of the deviation of the absolute frequency in Hz, and f_ {c} is a carrier frequency in Hz; determine (11-3) the deviation of the timing, T_ {off}, during a sleep period of one given duration Tsleep, according to 25 determines (11-4) the size of half the width of a search window assuming a maximum error of the sample clock of the receiver of agreement with T_ {wind / 2} = T_ {sleep} \ cdot t_ {ores} , assuming a maximum error, t_ {ores}, of the sample clock, t_ {0}; counting from the time t = 0 when the sleep mode and determining that the receiver should sleep (11-5) until time t expressed by 26 and look for (11-6) the start of the plot until the time expressed by 27 2. The apparatus of claim 1, wherein the timing correction unit (100) performs the offset of timing deviation using a frequency estimation and a channel estimate in the domain of the frequency 3. The apparatus of claim 2, which it also comprises a demodulation section (50) comprising the demodulator and timing correction unit (100), and in which the demodulation section also includes: a deviation estimation unit of frequency (102, 150) that provides an estimate of the frequency deviation, t0; a frequency correction unit (104) that receives the modulated baseband signal and provides a signal of frequency modulated baseband corrected; a Fast Fourier transform (FFT) (110) which receives the frequency modulated baseband signal corrected and provides a modulated domain subcarrier signal of the frequency; a channel estimation unit (112) that uses the frequency modulated baseband signal corrected for generate a channel estimate in the frequency domain that is applied to the timing correction unit (100 (4)); where the timing unit generates a corrected time channel estimate; a demodulator 114 that uses the modulated signal of frequency domain and channel estimation in the domain of the corrected time frequency to generate a signal demodulated 4. The apparatus of claim 3, wherein the timing correction unit (100 (4)) estimates a value of the timing deviation and compensates for the value of the timing deviation (120, t_ {off}) in the domain of the frequency by applying an appropriate phase factor to the subcarrier to update the channel estimate and therefore provide a Corrected time channel estimate. 5. The apparatus of claim 4, wherein the timing correction unit updates the estimate of channel using a relationship 28 where: H_ {m} [k ] is the corrected time channel estimate for a time index measured in data symbols k; H_ {m} [0 ] is the channel estimate in the frequency domain for the data symbol k; \ varphi_ {m, k} is the phase factor; and in which m is a subcarrier index for the subcarriers used. 6. The apparatus of claim 5, wherein the phase factor \ varphi_ {m, k} is defined by the following expression: 29 where: m is the index of the subcarrier; k is the time index measured in symbols of data; T_ {s} is a time symbol; T is a sample time; T init is a time between a reference time and a first data symbol; Y t_ {0} is the value of the timing deviation. 7. The apparatus of claim 3, wherein the unit of frequency deviation estimation is a unit of estimation of the frequency deviation directed by preamble (102) receiving the modulated baseband signal. 8. The apparatus of claim 7, which it also includes a deviation estimation unit of decision driven frequency (150) that is connected to receive respective inputs from the demodulation unit, the timing correction unit, and the transform of Fast Fourier (FFT). 9. The apparatus of claim 3, wherein the unit of frequency deviation estimation is a unit of decision-driven frequency deviation estimate (150) which is connected to receive respective inputs from the unit demodulation, timing correction unit, and the Fast Fourier transform (FFT). 10. The apparatus of claim 3, wherein the update of the channel estimate for the compensation of the timing deviation takes place every M avo symbol. 11. The apparatus of claim 3, wherein a value for M is selected based on an adaptation mode of particular link. 12. The apparatus of claim 1, which it also comprises a demodulation section (50) comprising the timing correction unit (100), and in which the section Demodulation includes: a deviation estimation unit of frequency (102, 150) that provides an estimate of the frequency deviation, t0; a frequency correction unit (104) that receives the modulated baseband signal and provides a signal of frequency modulated baseband corrected; a Fast Fourier transform (FFT) (110) which receives the digital complex modulated baseband signal from frequency corrected and provides, for each subcarrier, a modulated subcarrier signal of the frequency domain which is applied to the demodulator; a channel estimation unit (112) that uses the frequency band modulated baseband signal corrected and generates a channel estimate in the frequency domain; in which the timing unit receives the estimation of frequency deviation and subcarrier signal frequency domain modulation to generate a signal time frequency domain modulated subcarrier corrected: a demodulator (114) that uses the modulated signal of frequency domain and channel estimation in the domain of the frequency to generate a demodulated signal. 13. The apparatus of claim 12, in the that the timing correction unit estimates a value of the timing deviation and compensates for the deviation value of timing in the frequency domain by applying a appropriate phase factor to a subcarrier to update the signal frequency domain modulated subcarrier and provide therefore a modulated signal of the time frequency domain corrected. 14. The apparatus of claim 13, in the that the timing correction unit updates the signal frequency domain modulation using the relationship 30 where: R_ {TD, m} [k] is the modulated signal of the corrected time frequency domain of a subcarrier m ava of a k avo symbol that carries data; R_ {FFT, m} [k] is the frequency domain modulated signal as it is extracted by the Fast Fourier transform (FFT) of the subcarrier m ava of the k avo symbol that carries data; and m m, k is the phase factor; and in which m is a subcarrier index for the subcarriers used. 15. The apparatus of claim 14, in the that the phase factor \ varphi_ {m, k} is defined by the following expression: 31 where: m is the index of the subcarrier; k is the time index measured in symbols of data; T_ {s} is a time symbol; T is a sample time; T_ {init} is a time between a time of reference and a first data symbol; Y t_ {0} is the value of the timing deviation. 16. The apparatus of claim 12, in the that the unit of frequency deviation estimation is a unit of frequency deviation estimation directed by preamble (102) that receives the modulated baseband signal. 17. The apparatus of claim 16, which it also includes a deviation estimation unit of decision driven frequency (150) that is connected to receive respective inputs from the demodulation unit, the timing correction unit, and the estimation unit of channel. 18. The apparatus of claim 12, in the that the unit of frequency deviation estimation is a unit of frequency deviation estimation directed by decision (150) that is connected to receive respective inputs from the demodulation unit, the correction unit of timing, and channel estimation unit. 19. The apparatus of claim 1, wherein the timing correction unit uses a deviation from the frequency to determine a deviation value of the timing, in which the subcarrier signal comprises a data symbol stream, which also includes: a deviation estimation unit of frequency (150) that calculates: an estimated phase deviation for each data symbol as a function of the data symbol; a predicted phase deviation for each data symbol as a function of the estimated phase deviation of the same and an estimated phase deviation from a preceding one of the data symbols in the stream; a predicted sample phase deviation for each data symbol as a function of a phase deviation prediction of a corresponding one of the data symbols; Y the frequency deviation as a function of the predicted sample phase deviation for each signal sample of data. 20. The apparatus of claim 19, in the that the frequency deviation calculation unit comprises: a phase locked loop (168, 170) for generate the predicted phase deviation; a phase discrimination unit (166) for generate an estimated phase deviation for each data signal as a function of it; a filter (168) to receive deviations from Estimated phase and generate the predicted phase deviation for each data symbol as a function of the estimated phase deviation of the same and the estimated phase deviation of a preceding one of the Data symbols 21. The apparatus of any of the preceding claims, wherein the different carriers of Modulated radio frequencies have been modulated using Orthogonal Frequency Division Multiplexing (OFDM).